Ι have problem when type :
import tensorflow as tf
mnist = tf.keras.datasets.mnist
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0
model = tf.keras.models.Sequential([
tf.keras.layers.Flatten(input_shape=(28, 28)),
tf.keras.layers.Dense(128, activation='relu'),
tf.keras.layers.Dropout(0.2),
tf.keras.layers.Dense(10, activation='softmax')])
model.compile(optimizer='adam',
loss='sparse_categorical_crossentropy',
metrics=['accuracy'])
model.fit(x_train, y_train, epochs=5)
model.evaluate(x_test, y_test)
The result should be like this : Using Google Collab
But when I try using Jupyter notebook the result become like this: Jupyter ss1 Jupyter ss2jupyter 3jupyter4
CodePudding user response:
If I'm interpreting this correctly, you want to produce the exact loss and accuracy of the other model shown. Just FYI, this doesn't mean there is a problem with your model (evidently by your loss and accuracy values it is working rather well); no two models are likely to yield the same results, due to the optimising process of generating random weights and optimising them over a given number of iterations using your desired method.
As a previous answer has said, you can set a seed to reproduce the same results every time, using tf.random.set_seed(<your_seed>)
. However, the original model you have included does not not include this line, so you won't be able to reproduce those exactly.